The light environment in water bodies changes with depth due to the absorption of short and long wavelengths. Below 10 m depth, red wavelengths are almost completely absent rendering any red-reflecting animal dark and achromatic. However, fluorescence may produce red coloration even when red light is not available for reflection. A large number of marine taxa including over 270 fish species are known to produce red fluorescence, yet it is unclear under which natural light environment fluorescence contributes perceptively to their colours. To address this question we: (i) characterized the visual system of Tripterygion delaisi, which possesses fluorescent irides, (ii) separated the colour of the irides into its reflectance and fluorescence components and (iii) combined these data with field measurements of the ambient light environment to calculate depth-dependent perceptual chromatic and achromatic contrasts using visual modelling. We found that triplefins have cones with at least three different spectral sensitivities, including differences between the two members of the double cones, giving them the potential for trichromatic colour vision. We also show that fluorescence contributes increasingly to the radiance of the irides with increasing depth. Our results support the potential functionality of red fluorescence, including communicative roles such as species and sex identity, and non-communicative roles such as camouflage.

BackgroundOrganisms adapt to fluctuations or gradients in their environment by means of genetic change or phenotypic plasticity. Consistent adaptation across small spatial scales measured in meters, however, has rarely been reported. We recently found significant variation in fluorescence brightness in six benthic marine fish species across a 15 m depth gradient. Here, we investigate whether this can be explained by phenotypic plasticity alone, using the triplefin Tripterygion delaisi as a model species. In two separate experiments, we measure change in red fluorescent brightness to spectral composition and ambient brightness, two central parameters of the visual environment that change rapidly with depth.ResultsChanging the ambient spectra simulating light at −5 or −20 m depth generated no detectable changes in mean fluorescence brightness after 4–6 weeks. In contrast, a reduction in ambient brightness generated a significant and reversible increase in mean fluorescence, most of this within the first week. Although individuals can quickly up- and down-regulate their fluorescence around this mean value using melanosome aggregation and dispersal, we demonstrate that this range around the mean remained unaffected by either treatment.ConclusionWe show that the positive association between fluorescence and depth observed in the field can be fully explained by ambient light brightness, with no detectable additional effect of spectral composition. We propose that this change is achieved by adjusting the ratio of melanophores and fluorescent iridophores in the iris.Electronic supplementary materialThe online version of this article (doi:10.1186/s13104-016-1911-z) contains supplementary material, which is available to authorized users.

Active sensing has been well documented in animals that use echolocation and electrolocation. Active photolocation, or active sensing using light, has received much less attention, and only in bioluminescent nocturnal species. However, evidence has suggested the diurnal triplefin Tripterygion delaisi uses controlled iris radiance, termed ocular sparks, for prey detection. While this form of diurnal active photolocation was behaviourally described, a study exploring the physical process would provide compelling support for this mechanism. In this paper, we investigate the conditions under which diurnal active photolocation could assist T. delaisi in detecting potential prey. In the field, we sampled gammarids (genus Cheirocratus) and characterized the spectral properties of their eyes, which possess strong directional reflectors. In the laboratory, we quantified ocular sparks size and their angle-dependent radiance. Combined with environmental light measurements and known properties of the visual system of T. delaisi, we modeled diurnal active photolocation under various scenarios. Our results corroborate that diurnal active photolocation should help T. delaisi detect gammarids at distances relevant to foraging, 4.5 cm under favourable conditions and up to 2.5 cm under average conditions. Because ocular sparks are widespread across fish species, diurnal active photolocation for micro-prey may be a common predation strategy.

Ocular reflectors enhance eye sensitivity in dim light, but can produce reflected eyeshine when illuminated. Some fish can occlude their reflectors during the day. The opposite is observed in cryptic sit-and-wait predators such as scorpionfish and toadfish, where reflectors are occluded at night and exposed during the day. This results in daytime eyeshine, proposed to enhance pupil camouflage by reducing the contrast between the otherwise dark pupil and the surrounding tissue. In this study, we test this hypothesis in the scorpionfish Scorpaena porcus and show that eyeshine is the result of two mechanisms: the previously described Stratum Argenteum Reflected (SAR) eyeshine, and Pigment Epithelium Transmitted (PET) eyeshine, a newly described mechanism for this species. We confirm that the ocular reflector is exposed only when the eye is light-adapted, and present field measurements to show that eyeshine reduces pupil contrast against the iris. We then estimate the relative contribution of SAR and PET eyeshine to pupil brightness. Visual models for different light scenarios in the field show that daytime eyeshine enhances pupil camouflage from the perspective of a prey fish. We propose that the reversed occlusion mechanism of some cryptobenthic predators has evolved as a compromise between camouflage and vision.

BackgroundNatural red fluorescence is particularly conspicuous in the eyes of some small, benthic, predatory fishes. Fluorescence also increases in relative efficiency with increasing depth, which has generated speculation about its possible function as a “light organ” to detect cryptic organisms under bluish light. Here we investigate whether foraging success is improved under ambient conditions that make red fluorescence stand out more, using the triplefin Tripterygion delaisi as a model system. We repeatedly presented 10 copepods to individual fish (n = 40) kept under a narrow blue-green spectrum and compared their performance with that under a broad spectrum with the same overall brightness. The experiment was repeated for two levels of brightness, a shaded one representing 0.4% of the light present at the surface and a heavily shaded one with about 0.01% of the surface brightness.ResultsFish were 7% more successful at catching copepods under the narrow, fluorescence-friendly spectrum than under the broad spectrum. However, this effect was significant under the heavily shaded light treatment only.ConclusionsThis outcome corroborates previous predictions that fluorescence may be an adaptation to blue-green, heavily shaded environments, which coincides with the opportunistic biology of this species that lives in the transition zone between exposed and heavily shaded microhabitats.Electronic supplementary materialThe online version of this article (doi:10.1186/s12898-017-0127-y) contains supplementary material, which is available to authorized users.

Cases where animals use controlled illumination to improve vision are rare and thus far limited to chemiluminescence, which only functions in darkness. This constraint was recently relaxed by studies on Tripterygion delaisi , a small triplefin that redirects sunlight instead. By reflecting light sideways with its iris, it has been suggested to induce and detect eyeshine in nearby micro-prey. Here, we test whether ‘diurnal active photolocation’ also improves T. delaisi 's ability to detect the cryptobenthic sit-and-wait predator Scorpaena porcus, a scorpionfish with strong daytime retroreflective eyeshine. Three independent experiments revealed that triplefins in which light redirection was artificially suppressed approached scorpionfish significantly closer than two control treatments before moving away to a safer distance. Visual modelling confirmed that ocular light redirection by a triplefin is sufficiently strong to generate a luminance increase in scorpionfish eyeshine that can be perceived by the triplefin over 6–8 cm under average conditions. These distances coincide well with the closest approaches observed. We conclude that light redirection by small, diurnal fish significantly contributes to their ability to visually detect cryptic predators, strongly widening the conditions under which active sensing with light is feasible. We discuss the consequences for fish eye evolution.

Since the discovery of red fluorescence in fish, much effort has been invested to elucidate its potential functions, one of them being signaling. This implies that the combination of red fluorescence and reflection should generate a visible contrast against the background. Here, we present in vivo iris radiance measurements of Tripterygion delaisi under natural light conditions at 5 and 20 m depth. We also measured substrate radiance of shaded and exposed foraging sites at those depths. To assess the visual contrast of the red iris against these substrates, we used the receptor noise model for chromatic contrasts and Michelson contrast for achromatic calculations. At 20 m depth, T. delaisi iris radiance generated strong achromatic contrasts against substrate radiance, regardless of exposure, and despite substrate fluorescence. Given that downwelling light above 600 nm is negligible at this depth, we can attribute this effect to iris fluorescence. Contrasts were weaker in 5 m. Yet, the pooled radiance caused by red reflection and fluorescence still exceeded substrate radiance for all substrates under shaded conditions and all but Jania rubens and Padina pavonia under exposed conditions. Due to the negative effects of anesthesia on iris fluorescence, these estimates are conservative. We conclude that the requirements to create visual brightness contrasts are fulfilled for a wide range of conditions in the natural environment of T. delaisi.

Active sensing has been well documented in animals that use echolocation and electrolocation. Active photolocation, or active sensing using light, has received much less attention, and only in bioluminescent nocturnal species. However, evidence has suggested the diurnal triplefin Tripterygion delaisi uses controlled iris radiance, termed ocular sparks, for prey detection. While this form of diurnal active photolocation was behaviourally described, a study exploring the physical process would provide compelling support for this mechanism. In this paper, we investigate the conditions under which diurnal active photolocation could assist T. delaisi in detecting potential prey. In the field, we sampled gammarids (genus Cheirocratu s) and characterized the spectral properties of their eyes, which possess strong directional reflectors. In the laboratory, we quantified ocular sparks size and their angle-dependent radiance. Combined with environmental light measurements and known properties of the visual system of T. delaisi , we modeled diurnal active photolocation under various scenarios. Our results corroborate that diurnal active photolocation should help T. delaisi detect gammarids at distances relevant to foraging, 4.5 cm under favourable conditions and up to 2.5 cm under average conditions. To determine the prevalence of diurnal active photolocation for micro-prey, we encourage further theoretical and empirical work.